Fully decoupled neural network learning using delayed gradients

Training neural networks with back-propagation (BP) requires a sequential passing of activations and gradients. This has been recognized as the lockings (i.e., the forward, backward, and update lockings) among modules (each module contains a stack of layers) inherited from the BP. In this paper,...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhuang, Huiping, Wang, Yi, Liu, Qinglai, Lin, Zhiping
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2024
Subjects:
Online Access:https://hdl.handle.net/10356/174476
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English